SU-D-WAB-04: Restoration of CBCT Images by Deformable Registration for Plan Evaluation and Replanning
Bibliographic record
Abstract
Purpose: To use a simple deformable image registration (DIR) scheme to correct CBCT images in order to use them for treatment plan assessment and replanning. Methods: A dataset of five prostate patients with weekly CBCT scans was available for this study. Contours of soft-tissue organs were made on the planning CT as well as on each CBCT. The following procedure was applied: (1) a mask was automatically generated to exclude certain area from the registration (table couch, CBCT truncated region due to limited field of view); (2) a B-spline based DIR algorithm was used to register each plan CT to daily CBCTs; (3) CBCT regions where air appeared (i.e. air is seen on the CBCT but not on the plan CT) were automatically segmented and excluded from the registration because of the poor performance of DIR algorithms in such cases. Results: Prior to DIR, the average differences in Houndsfield units between CT and CBCT were 26, 7, 107, 125 for the prostate, bladder, seminal vesicles and rectum respectively. After DIR these differences were respectively reduced to 5, 5, 4, 2. The external contours of the patients were assessed using Dice's coefficient. On average the Dice coefficient between CT and weekly CBCTs was 0.92. After DIR this value was improved to 0.97. Furthermore by registering the larger CT to shorter CBCT (i.e. 64 slices only) we obtain realistic scattering media outside the immediate CBCT field of view. The whole process takes about 4 minutes although no explicit speed optimization has been done yet. Conclusion: This technique has several advantages for plan assessment and/or replanning purposes: it combines the accurate Houndsfield numbers from the plan CT with the daily anatomy of the CBCT, preserve the external contour and provide realistic scattering media outside the CBCT field of view.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".